The present invention relates to a method for controlling a process for the production of a cellulose fibre containing product from cellulose fibre containing raw material, in particular a process for the production of wood based panels such as particleboards.
The production of cellulose fibre containing products often involve addition of different substances during the production process. The purpose of the addition may be to facilitate the production process per se, or to effect the resulting product, or both. Addition of various glues, often consisting of a glue system comprising a resin and a hardener, in the production of wood based panels is exemplary of an addition in order to effect the resulting product. An important parameter in this connection is often the proportion between the components making up the substance, if more than one, such as e.g. the proportions between resin and hardener in a glue, or the proportions between urea and formaldehyde in a glue containing those components. Another important parameter may be the proportions between different added substances, which are destined to different parts of the panel, e.g. in different layers of the panel.
WO 97/04299 relates to a method for controlling process variables influencing parameters of wood based panels, including glue amounts and wax amounts, which method comprises analysing the raw material by means of a spectrometric method, linking the obtained spectral data into a combination with desired parameters, and comparing said combination with reference combinations consisting of reference data from reference raw material linked with known parameters of said reference material, whereby the reference combinations have been calibrated to known variables by means of multivariate analysis. There is however no suggestion about controlling any proportions relating to those additives.
It would however be desirable to be able to control such proportions, as by doing so, it should be possible to produce cellulose fibre containing product such as wood based panels with very specific properties at optimised technical and economical conditions.
The problem to be solved by the present invention is thus to enable such controlling.
This problem has been solved by the invention as defined by the appended claims. More particularly the present invention involves a method for controlling a process for the production of a cellulose fibre containing product from cellulose fibre containing raw material, during which process various substances are added, at least two of which substantially influence the same property of said product in correspondence to the relation between said substances, whereby a calibration model has been established by way of a procedure comprising linking known reference values for said product property and corresponding known reference relations between the substances, by means of a mathematical function; the method comprises the steps of
I) applying the calibration model on the actual relation between the substances in order to predict the value for said product property;
II) comparing the predicted value for said product property with a desired target value for said product property;
and, if said predicted value is not substantially equal to said target value, adjusting the actual relation between the substances in a predetermined way;
III) repeating steps I and II until said predicted value is substantially equal to said target value.
By the expression xe2x80x9cin a predetermined wayxe2x80x9d is, in the present context, partly meant that it has been determined in advance whether the relation should be adjusted at all in the present turn in the loop defined by steps I to III; partly meant that it has been determined in advance whether the adjustment should be incremental or decremental depending on the fact per se that the predicted value is larger that the target value, irrespective of the size of the difference between those values, and vice versa, and partly meant that the absolute value of the adjustment has been determined in advance.
The present method presents a plurality of advantages to the producer of cellulose fibre containing products. By means of the present method a particleboard producer, for instance, can regulate one or more properties of the produced boards in a very fast and economical way.
In a preferred embodiment the calibration model has been established by means of a multivariate analysis; this also implies that the same kind of multivariate analysis is applied when the calibration model is applied on the actual relation. The multivariate analysis may, for instance, be principal component analysis (PCA), partial least squares regression (PLS), principal component regression (PCR), multilinear regression analysis (MLR), discriminant analysis, or any other suitable method for multivariate analysis. The PLS method is disclosed in detail in Carlson R., xe2x80x9cDesign and optimization in organic synthesisxe2x80x9d, B. G. M. Vandeginste, O. M. Kvalheim, Eds., xe2x80x9cData handling in science and technologyxe2x80x9d, (Elsevier, 1992), vol. 8, incorporated herein by reference. For a tutorial in PCA, PLS and PCR, see P. Geladi et al in xe2x80x9cPartial Least-Squares Regression: A Tutorialxe2x80x9d in Anal. Chim. Acta, 185, 1-17 (1986), which is incorporated herein by reference in its entirety. By MLR, the best fitting plane for the parameters as a function of the spectra is defined, using least squares techniques to define each boundary of the plane. This plane is then used to recognize and assign a predicted value to an unknown parameter value. Discriminant Analysis is a method whereby, by use of spectral data, the known parameter values are grouped into different clusters, separated by linear decision boundaries. From its spectrum, a sample of unknown parameter values then can be matched to a cluster, and the parameter value can be assigned a value, e.g., the average value of the cluster. Applied generally to the field of chemistry those statistic methods are also termed chemometrics methods. The technique of chemometrics is more fully explained in S. D. Brown, xe2x80x9cChemometricsxe2x80x9d, Anal. Chem. 62, 84R-101R (1990), which by reference is incorporated herein in its entirety.
In a preferred embodiment the cellulose fibre containing product is a wood based panel. The wood based panel may comprise two or more layers. One of the substances may for instance be added to substantially stay between the layers, in order to e.g. agglutinate (i.e. glue together) the layers. Alternatively, one substance may be added to be substantially contained within one of the layers, in order to e.g. bind wood based particles in a particle board, or to provide the layer with some property, or to enhance some already present property. The wood based panel is preferably a board, such as a particleboard, a medium density fibre board (MDF), a waferboard, an oriented strand board (OSB), a hardboard, or a board of plywood; in particular, the board is a particleboard. Background information on particleboard and the processes for the manufacturing thereof is set forth in xe2x80x9cModern Particleboard and dry-process fibreboard manufacturingxe2x80x9d by Thomas M. Maloney (1993), (cf. especially Chapter 4 and 5), which by reference is incorporated herein in its entirety.
It should be understood that the two (or more) substances which have a substantial influence on the same property of said product in correspondence to the relation between said substances may very well lack any effect, or only have a weak effect on that (or any other) property when used isolated from each other. This, basically, is the case when, for instance, one substance is a resin and another one is a corresponding hardener. On the other hand, the effect of the substances may in another embodiment be of basically the same magnitude, e.g. when using two urea-formaldehyde glues having different U/F proportions. In the latter case the property of interest can be brought towards its target value by controlling the U/F proportion of an actually added, mixed glue within a range defined by the U/F proportions of the two glues, by applying the present method on the relation between the added and mixed amounts of the glues. If instead the glues are added to different layers of a board, and thus basically are not mixed with each other, the property of interest can still be controlled, and in this case the actual relation between the added, but not mixed, substances is used in the construction and application of the calibration model.
In one embodiment each one of the substances is a glue containing an amino resin, such as for instance a urea-formaldehyde resin (UF), a melamine-urea-formaldehyde resin (MUF), or a phenol- formaldehyde resin (PF), whereby the substances may differ from each other with regard to e.g. the content of formaldehyde. Other glues may however also be used, such as e.g. isocyanate resin (MDI).
The product properties that can be controlled by the present method are for instance density, density profile, internal bond strength, thickness swelling, absorption value, permeability value, perforator value, modulus of rapture (MOR), modulus of elasticity (MOE), parameters relating to volatile organic compounds (VOC), and emission chamber value; this, however, is not an exhaustive list of controllable properties. Basically any measurable property of the product may be controlled by the present method.
The total amount of substance is often also important with regard to the resulting product property. Thus, in one embodiment of the invention, in which the two substances have a substantial influence on said property also in correspondence to the added combined amount of the substances, the calibration model involves linking said known values for said product property with the corresponding known combined amounts of the substances added and the corresponding relations between the substances in said amounts; and the application of the calibration model involves linking of the actual relations between the substances and the corresponding relations between the substances in said amounts in order to predict the actual value for said product property.
In a particularly preferred embodiment of the present method the calibration model has been established by way of a procedure comprising analysing reference raw material, by means of an analysis method, to provide reference sample data, and linking said reference sample data with known values for said product property and corresponding known relations between the substances, by means of said mathematical function; the method additionally comprises the step of analysing the raw material, by means of said analysis method, to provide sample data; and the application of the calibration model involves linking of the sample data with the relation between the substances in order to predict the actual value for said product property.
This embodiment provides additional advantages to the producer of cellulose fibre containing products. By means of the present method a particleboard producer, for instance, may manufacture boards with high surface strength even from wooden raw material of inferior quality by compensating the lacking quality with a urea-formaldehyde glue having a high molar proportion of formaldehyde, which is used in the surface layer of the board, while the control of the dosage and the molar proportion of formaldehyde in the glue in the core layer based upon other desired properties such as for instance a specified density and/or a low formaldehyde emission value.
The sample data is preferably transferred into latent variables before being linked with the target value, as are preferably the reference sample data before they are linked with said known values.
In a particularly preferred embodiment of the present invention the analysis method is a spectrometric method. The sample data as well as the references sample data is preferably processed in this instance to reduce noise and is also suitably adjusted for drift and diffuse light scatter, for instance by means of the Kubelka-Munk transformation (P. Kubelka, F. Munk, Z. Tech. Physik 12, 593 (1931), incorporated herein by reference), which takes account of absorption and scatter, the Multiplicative Scatter Correction (P. Geladi, D. MacDougall, H. Martens, Appl. Spect. 39, 491-500 (1985), incorporated herein by reference) where each spectrum is xe2x80x9ccorrectedxe2x80x9d in both offset and slope by comparing it to an xe2x80x9cidealxe2x80x9d spectrum (the mean spectrum). Another way of linearizing the spectral data involves the use of derivatives, e.g. up to the fourth order derivatives (A. Savitzky, M. J. E. Golay, Anal. Chem. 36, 1627-1639 (1964), incorporated herein by reference). The derivative of the spectrum results in a transformed spectrum, consisting only of the relative changes between the adjacent wavelengths, and it has been shown that the peak intensities of derived spectra tend to be more linear with concentration (T. C. O""Haver, T. Begley, Anal. Chem. 53, 1876 (1981), incorporated herein by reference). Linearization can also be accomplished by use of the Fourier transformation, or by use of the Standard Normal Variate transformation (R. J. Barnes, M. S. Dhanoa and S. J Lister, Appl. Spectrosc., Vol. 43, number 5, pp. 772-777 (1989), incorporated herein by reference). This noise reduction/drift and diffuse light scatter-adjustment is suitably done before the data is linked as set forth above. The spectrometric method may be an absorption, reflectance, emission or transmission spectrometric method, or any other conceivable spectrometric method. Although the spectrometric method may relate to any suitable kind of radiation at any wavelength range, it is preferred to use a spectrometric method operating in the wavelength range from about 180 to about 2500 nm, particularly from about 400 to 2500 nm, and especially from about 1000 nm to about 2500 nm. It is particularly preferred to operate the present method in the near-infrared radiation (NIR) range. The principles of NIR spectroscopy are described by Williams, P.; Norris, K. (1987): Near-Infrared Technology in the Agriculture and Food Industries, AACC, St. Paul/Min. and Stark, E.; Luchter, K. (1986): Near Infrared Analyses (NIRA) A Technology for Quantitative and Qualitative Analyses. (Applied Spectroscopy Revues 22:4), all of which are hereby incorporated by reference.
Technically, the spectrometric analysis can be performed by on-line, in-line or at-line optical fibre probing, or by taking individual samples for separate analysis. In any case the spectra are preferably subjected to further data treatment using values from several discrete wavelengths from each particular spectrum. The radiation used in the spectrometric method preferably impinges directly on the raw material.
In a specifically preferred embodiment of the present invention the produced cellulose fibre containing product is a particleboard comprising one core layer and two surface layers; the cellulose fibre contains raw material comprises sawdust, shavings, chips or shavings from round wood, or a combination thereof; the controlled product property is density, density profile, internal bond strength, MOR, MOE, thickness swelling, absorption value, permeability value, perforator value, or emission chamber value; the substances are glues obtained by mixing formaldehyde, urea, and optionally any other suitable component; and the glues differ from each other with regard to the proportions between formaldehyde and urea, whereby a first glue holds a higher proportion between formaldehyde and urea than a second glue.
In another embodiment the present method for controlling a process for the production of a cellulose fibre containing product is combined with EP 564,013, which relates to a method and a device for mixing binders using at least two components which are flowed from separate storage containers towards a common mixing point. The storage containers are continuously weighed, and their change of weight per unit of time is determined.
The present invention will now be further illustrated by means of some non-limiting examples.